To reduce the bandwidth required for transferring video data as much as possible, multiple video compression methods may be selected to compress video data. The video compression methods include: intra-frame compression and inter-frame compression. Currently an inter-frame compression method based on motion estimation is often selected. Specifically, the process that an inter-frame compression method is selected at an image encoder to compress and encode an image includes: an encoder partitions an image block to be encoded into several sub-image blocks of equal sizes, for each sub-image block, searches a reference image for an image block that matches with the current sub-image block most as a prediction block, performs subtraction on corresponding pixel values of the sub-image block and the prediction block to obtain a residual, performs entropy encoding on a value obtained after transform and quantization of the residual, and in the end sends to a decoder a bitstream obtained through entropy encoding and motion vector information together. The motion vector information indicates a position difference between the current sub-image block and the prediction block. At the decoder for an image, an entropy encoded bitstream is first acquired to perform entropy decoding to obtain a corresponding residual and corresponding motion vector information, a corresponding matching image block (namely, the prediction block) is then acquired from the reference image according to the motion vector information, and the values of corresponding pixel points in the image block and the residual value are added to obtain a current sub-image block.
During the process, an encoder needs to partition an image block to be encoded into some sub-image blocks. For example, a horizontal partition manner or a vertical partition manner may be selected to partition an image block into two rectangular sub-image blocks. Alternatively, a partition manner from lower left to upper right or a partition manner from lower right to upper left may be selected to partition an image block into two sub-image blocks, in which one sub-image block is a triangle. In the prior art, two parameters are selected for identification, in which a numerical value such as 1, 2, 3, 4, . . . , is selected for one parameter to identify a partition manner, for example, 1 identifies horizontal partition, 2 identifies vertical partition, 3 identifies a partition manner at 45 degrees from lower left to upper right, 4 identifies a partition manner at 45 degrees from lower right to upper left, 5 identifies a partition manner at 30 degrees from lower left to upper right, 6 identifies a partition manner at 30 degrees from lower right to upper left, and the like; the other parameter identifies the position of a sub-image block endpoint.
The prior art has the following disadvantages.
In the two parameters selected in the prior art, the first parameter identifies a partition manner, and as at least 4 partition manners (horizontal partition, vertical partition, partition at 45 degrees from lower left to upper right, and partition at 45 degrees from lower right to upper left) need to be identified, the parameter needs at least two bits. The bit count of the parameter for identifying a sub-image block obtained by partitioning an image block is the sum of the bit count of the first parameter and the bit count of the second parameter, so that the number of bits occupied thereof is large.
The first parameter identifies a partition manner. A different numerical value is needed to identify a partition at a different angle, for example, 4 identifies the partition manner at 45 degrees from lower right to upper left, 5 identifies a partition manner at 30 degrees from lower left to upper right, 6 identifies a partition manner at 30 degrees from lower right to upper left, and the like. Generally speaking, the first parameter does not select many bits, so that only a limited number of angle values are capable of being identified, resulting in limited types of partition when a partition manner from lower right to upper left or from lower left to upper right is selected.